Survival status and predictors of mortality among patients with cervical cancer in two oncology units in Northwest Ethiopia: a retrospective follow-up study Authors: Aragaw Tesfaw, Adamu Addissie, Eva Johanna Kantelhardt, Sefonias Getachew Published: 2025 ABSTRACT AbstractBackgroundCervical cancer is the most frequently diagnosed malignancy and a leading cause of death among women in sub-Saharan Africa, including Ethiopia. Despite its high mortality rate, little is known about survival outcomes and predictors of mortality for patients with cervical cancer in Northwest Ethiopia.MethodsA hospital based retrospective follow-up study was conducted among patients with cervical cancer diagnosed and treated from 1 February 2017 to 30 January 2022 at Felege Hiwot Hospital and University of Gondar Hospitals. Data were collected using a data abstraction checklist and telephone interviews. The Kaplan–Meier and the log-rank test were used to estimate and compare survival probabilities. Sensitivity analysis was performed assuming that patients lost to follow-up may have died 6 months after their last visit. The Cox regression model identified predictors of mortality with 95% confidence intervals, considering p < 0.05 as statistically significant.ResultsThe overall survival rate at 52 months of follow-up was 18.63%. The survival rates at 2, 3, and 4 years were 81.2%, 55.89%, and 37.26%, respectively. In the worst-case analysis, survival at 2, 3, and 4 years declined to 65%, 39%, and 26%, respectively. Significant predictors of mortality included age ≥ 50 years (AHR = 3.4, 95% CI: 1.80–6.48, p < 0.0001), HIV positivity (AHR = 2.9, 95% CI: 1.55–5.61, p = 0.001), Travel distance > 65 km (AHR = 1.9, 95% CI: 1.28–3.08, p = 0.002), no history of surgery (AHR = 2.8, 95% CI: 1.33–5.71, p = 0.006), and advanced stage at diagnosis ; Stage III: (AHR = 3.4, 95% CI: 1.95–5.79, p < 0.001), Stage IV (AHR = 6.7, 95% CI: 3.35–13.58, p < 0.001). However, residence, comorbidity, patient delay, anemia and menopausal status were not significant predictors of mortality (P > 0.05).ConclusionThe overall survival rate for patients with cervical cancer was low in this study. The identified predictors highlight the need to expand cervical cancer screening, diagnostic, and treatment services, especially in regions lacking radiotherapy facilities. Special focus should also be placed on elderly women, and women living with HIV.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12885-025-15444-7.Keywords: Cervical cancer, Survival, Predictors, Ethiopia Background In many low-income countries, infectious diseases remain the leading causes of death. However, there has been a significant rise in the incidence and mortality of non-communicable diseases, including cancer, creating a “double burden” that strains existing health care systems [1–3]. According to GLOBOCAN 2018 estimates, there were 18.1 million new cancer cases globally, with cervical cancer ranking fourth in both incidence and mortality [2]. The incidence of cervical cancer varies widely across regions, with age-standardized rates ranging from 4.4 to 75.9 per 100,000 people [4]. In sub-Saharan Africa, cervical cancer accounts for 23.3% of all cancer cases in women, with eastern Africa particularly affected—reporting an incidence of 40 cases per 100,000 women per year [2, 5]. If left unchecked, the annual number of cervical cancer cases is projected to rise from 570,000 to 700,000 by 2030, with deaths increasing from 311,000 to 400,000—most of them among women in low- and middle-income countries [2, 6]. Persistent infection with carcinogenic types of human papillomavirus is the main cause of cervical cancer, although it has also created new opportunities for primary and secondary prevention. Factors such as tobacco smoking, high parity, long-term use of hormonal contraceptives, and co-infection with human immunodeficiency virus (HIV) are recognized cofactors in disease progression [7, 8]. Approximately 87% of all cervical cancer deaths occur in sub-Saharan African countries, primarily due to advanced-stage presentation and limited access to screening and treatment. Nearly all of these fatalities could be prevented through screening, early detection of precancerous cervical lesions, and access to treatment for all women [5, 9]. Evidence from 2019 showed that cervical cancer has the highest age-standardized incidence rate in Ethiopia, at 24.6 per 100,000 population [10]. Each year, 7,445 women are newly diagnosed, making it the second most common cancer and a leading cause of cancer-related death among women in the country. The peak incidence occurs in women aged 40–49 years [11, 12]. According to the GLOBOCAN 2022 estimates, cervical cancer represents 8168 new cases, comprising 15.5% of all cancers among women in the country [13]. Although there are regional variations, several factors have been identified as predictors of mortality in patients with cervical cancer. Long-term survival also varies based on sociocultural and infrastructure differences in diagnosis and treatment. Tumor grade, regional organ involvement and recurrence, hemoglobin level, stage at presentation, tumor size, radiotherapy dose, and use of concurrent chemo-radiotherapy are among the factors that determine survival in patients with cervical cancer [14–16]. Because cervical cancer is the most preventable and curable cancer if treated early, focusing on prevention and early detection is vital for resource-limited countries like Ethiopia. The country has launched a national cancer control plan as part of its Health Sector Transformation Plan to address the problem [17]. In line with the global cervical cancer elimination initiative, Ethiopia’s Ministry of Health is working on vaccination, screening, and treatment of precancerous cervical lesions to eliminate the disease [18]. However, the country has limited cancer diagnostic and treatment centers, causing most patients to face long delays and present at advanced, incurable stages. Radiotherapy centers are few and inaccessible to many patients from different parts of the country [19, 20]. Despite the high burden of morbidity and mortality reported in the national cancer registry and hospital data, little is known about the survival status and predictors of mortality from cervical cancer in Northwest Ethiopia. We aimed to fill this gap because investigating survival at the local level is crucial for informing public health strategies and evaluating the effectiveness of cancer care—both in terms of service delivery and public awareness of available treatment options. Materials and methods Study settings, design, and population This study was conducted in the two oncology units of Felege Hiwot and University of Gondar comprehensive specialized hospitals in Northwest Ethiopia. Felege Hiwot Comprehensive Specialized Hospital (FHCSH) is located in Bahir Dar, approximately 565 km from Addis Ababa, while the University of Gondar Hospital (UOG) is about 737 km from the capital. Both serve as teaching hospitals for medical and health science students across various specialties. Each has a dedicated oncology unit providing diagnostic services, surgery, chemotherapy, and palliative care for patients with cancer referred from health centers, district hospitals, and general hospitals in the region. Chemotherapy, often combined with surgery, is one of the treatment modalities available. Because neither hospital has a radiotherapy machine, patients who can afford it are referred to the country’s only radiotherapy center, Tikur Anbesa Specialized Hospital. A retrospective follow-up study was conducted among patients with cervical cancer diagnosed and treated at the two oncology units between 1 February 2017 and 30 January 2022. Medical records of patients with pathologically confirmed cervical cancer and clear staging information were included, while unavailable records, incomplete charts, and patients diagnosed during the data collection period were excluded. Sampling and data collection procedure The medical records of patients with cervical cancer diagnosed between 1 February 2017 and 30 January 2022, who were registered and treated at the two oncology units, were reviewed. First, the medical registration numbers of women with cervical cancer were extracted from the chemotherapy follow-up logbook. All available patient charts were then located using the medical registration numbers, eligibility was assessed, and information was extracted based on the inclusion criteria. In total, 592 patients with cervical cancer were identified from the chemotherapy logbooks, of which 441 charts met the inclusion criteria and were included in the final analysis (Fig. 1). Flowchart showing the sampling procedure for patients with cervical cancer on follow-up at the oncology units of FHCSH and UOG, Northwest Ethiopia Data were collected from the patients’ medical records using a data extraction checklist and telephone calls. The checklist was developed after reviewing relevant literature from similar studies and other sources aligned with the study objectives [19–22]. It covered socio-demographic information; clinical, reproductive, and tumor characteristics; and behavioral factors (see supplementary file 1). Two data collectors and one supervisor were assigned at each hospital to gather the data. Chemotherapy registration books (standard patient follow-up), laboratory reports, and physical examination findings were extracted from the patients’ charts. The outcome status was determined through phone calls and information from medical records. Measurements The primary outcome of the study was time to death, defined as the period from the date of pathological diagnosis of cervical cancer to the end of follow-up (diagnosis between 1 February 2017 and 30 January 2022), and measured in months. Patients who were alive at the end of the study, lost to follow-up, or transferred to other care units were considered censored. The event was defined as the death of a patient, regardless of cause. Patients lost to follow-up for ≥ 6 months were classified as lost to follow-up [19]. While those lost for less than 6 months were considered alive unless death was confirmed through records or phone calls. Stage at diagnosis was determined using the revised FIGO staging for cervical cancer [23] and classified as early stage (I and II) or late stage (III and IV) [24, 25]. Hemoglobin levels below 12.0 g/dL were classified as anemic [26]. Patients diagnosed with hypertension, diabetes mellitus, asthma, or heart disease, as identified by the Charlson comorbidity index [27] were considered to have comorbidities. Patient delay (health-seeking interval) was defined as the time from first symptom recognition to the first visit to a health care facility, recorded from patient files as chief complaint, and categorized as short delay (≤ 3 months) or long delay (>3 months). Time to diagnosis (health system delay or diagnostic interval) was measured from the first health facility visit to confirmed diagnosis and categorized as short delay (≤ 30 days) or long delay (>30 days) [20, 28, 29]. Tumor metastasis was determined based on ultrasounds, chest X-rays, computed tomography scans, and clinical records. Distance was categorized using the average distance in kilometers. A worst-case analysis was performed by assuming all patients lost to follow-up (n = 83) experienced the event 6 months after their last follow-up date. Data analysis Data were analyzed using STATA version 16 software. Overall survival was estimated using the Kaplan–Meier curve, and a log-rank test was used to compare survival among groups with a 95% confidence interval (CI). The assumptions of the Cox proportional hazards regression model were checked using log (− log (St)) plots and a global test. Factors that met the proportional hazards assumption and had a p-value of < 0.05 in the bivariate Cox regression were included in the multivariate analysis. The Breslow test was used to handle tied failures. Multicollinearity and interaction in the main effects model were assessed, with variance inflation factors greater than 10 indicating multicollinearity. Model goodness-of-fit was evaluated using the Cox–Snell residual plot. The final model was presented using adjusted hazard ratios (AHRs) with corresponding 95% CIs, and a p-value of < 0.05 was used to determine statistical significance. Results Socio-demographic characteristics In total, 592 medical records of patients with cervical cancer were reviewed, and 441 complete records were included in the final analysis. The mean (± standard deviation) age of the patients was 49.13 ± 10.3 years. A total of 323 (73.2%) patients came from rural areas. Among the available records, nearly one-fourth [47 (20.3%)] of the patients were illiterate, and 42 (15.6%) were housewives. More than half [231 (62.9%)] were married. There was a significant difference in survival status across age groups (p < 0.0001), residence (p = 0.0042), and educational status (p = 0.002) (Table 1). Table 1Socio-demographic characteristics of patients with cervical cancer diagnosed at UOG and FHCSH, Northwest EthiopiaCharacteristicsStatus at last contactTotalLog-rank testp-valueCensoredDeathAge group ≤ 50 years235 (68.5%)27 (27.6%)262 (59.4%)60.37 < 0.0001 > 50 years108 (31.5%)71 (72.4%)179 (40.6%)Residence Rural239 (69.7%)84 (85.7%)323 (73.2%)8.18 0.0042 Urban104 (30.3%)14 (14.3%)118 (26.8%)Education (n = 231) Illiterate26 (15.1%)21 (35.6%)47 (20.3%)14.94 0.002 Primary education53 (30.8%)15 (25.4%)68 (29.4%) Secondary education66 (38.4%)16 (27.1%)82 (35.5%) College and above27 (15.7%)7 (11.9%)34 (14.7%)Religion (n = 372) Orthodox208 (71.0%)61 (77.2%)269 (72.3%)5.240.263 Muslim33 (11.3%)7 (8.9%)40 (10.8%) Protestant21 (7.2%)1 (1.3%)22 (5.9%) Catholic20 (6.8%)9 (11.4%)29 (7.8%) Other*11 (3.8%)1 (1.3%)12 (3.2%)Marital status (n = 367) Married186 (63.9%)45 (59.2%)231 (62.9%)0.830.975 Single15 (5.2%)3 (3.9%)18 (4.9%) Widowed49 (16.8%)15 (19.7%)64 (17.4%) Divorced24 (8.2%)7 (9.2%)31 (8.4%) Separated17 (5.9%)6 (7.9%)23 (6.3%)Occupation (n = 270) Farmer57 (25.8%)19 (38.8%)76 (28.1%)6.340.175 Government employee50 (22.6%)12 (24.5%)62 (23.0%) Housewife36 (16.13%)6 (12.2%)42 (15.6%) Other**45 (20.4%)8 (16.3%)53 (19.6%)Data are presented as n (%)*Traditional and Jehovah’s Witness**Student, private employee, daily laborerBoldface text: statistically significant Socio-demographic characteristics of patients with cervical cancer diagnosed at UOG and FHCSH, Northwest Ethiopia Data are presented as n (%) *Traditional and Jehovah’s Witness **Student, private employee, daily laborer Boldface text: statistically significant Clinical and reproductive characteristics As shown in Table 2, the majority of patients 288 (65.3%) were postmenopausal, and 48 (10.9%) were HIV-positive. Among the available records, 17 (4.1%) were pregnant at the time of diagnosis, and 38 (8.9%) had a family history of cervical cancer. A total of 190 (43.1%) patients had anemia, and 165 (37.4%) had comorbidities such as hypertension, diabetes, or heart disease. More than half [192 (62.3%)] had a history of contraceptive use. Traditional treatment use was reported in 21 (10.6%) patients. Cervical cancer was misdiagnosed in 80 (18.4%) patients at their first visit. Vaginal bleeding was the most common presenting complaint, reported by 418 (94.8%). A total of 336 (76.2%) patients experienced a long patient delay before seeking care, while 194 (44%) faced a long diagnostic delay within the health care system. Survival status differed significantly by menopausal status (p < 0.001), history of precancerous cervical lesions (p < 0.001), anemia (p = 0.04), comorbidities (p = 0.001), use of traditional treatments (p = 0.002), smoking (p = 0.028), alcohol use (p = 0.008), and patient delay (p = 0.006) according to the log-rank test (Table 2). Table 2Clinical characteristics of patients with cervical cancer diagnosed at UOG and FHCSH, Northwest EthiopiaCharacteristicsStatus at last contactTotalLog-rank testp-valueCensoredDeathMenopausal status Pre-menopausal136 (39.7%)17 (17.3%)153 (34.7%)25.01< 0.001* Menopausal207 (60.3%)81 (82.7%)288 (65.3%)HIV status Negative309 (90.1%)84 (84.7%)393 (89.1%)1.880.169 Positive34 (9.9%)14 (14.3%)48 (10.9%)Pregnancy status (n = 418) Not pregnant13 (4.0%)4 (4.2%)17 (4.1%)0.1030.748 Pregnant310 (96.0%)91 (95.8%)401 (95.9%)Family history of cervical cancer (n = 425) Yes28 (8.4%)10 (10.8%)38 (8.9%)0.1090.742 No304 (91.6%)83 (89.2%)387 (91.1%)History of precancerous cervical lesion (n = 416) Yes24 (7.2%)26 (31.3%)50 (12%)18.83< 0.001* No309 (92.8%)57 (68.7%)366 (88%)Anemia Yes136 (39.7%)54 (55.1%)190 (43.1%)4.220.04* No207 (60.3%)44 (44.9%)251 (56.9%)Comorbidities Yes117 (34.1%)48 (49.0%)165 (37.4%)11.120.001* No226 (65.9%)50 (51.0%)276 (62.6%)History of giving birth (n = 403) Yes228 (72.6%)67 (75.3%)295 (73.2%)0.1430.705 No86 (27.4%)22 (24.7%)108 (26.8%)Contraceptive use (n = 308) Yes151 (62.1%)41 (63.1%)192 (62.3%)0.0010.984 No92 (37.9%)24 (36.9%)116 (37.7%)Traditional treatment (n = 198) Yes10 (6.8%)111 (22%)21 (10.6%)9.990.002* No138 (93.2%)39 (78.0%)177 (89.4%)Smoking (n = 337) Yes7 (2.7%)3 (3.8%)10 (3.0%)7.160.028* No251 (97.3%)76 (96.2%)327 (97.0%)History of alcohol intake (n = 371) Yes20 (6.9%)12 (14.8%)32 (8.6%)7.030.008* No270 (93.1%)69 (85.2%)339 (91.4%)Diagnosis at first visit (n = 435) Misdiagnosed at first visit59 (17.5%)21 (21.6%)80 (18.4%)2.250.324 Diagnosed with cervical cancer95 (28.1%)31 (32.0%)126 (29.0%) No diagnosis made184 (54.4%)45 (46.4%)229 (52.6%)Patient delay interval Short (< 3 months)91 (26.5%)14 (14.3%)105 (23.8%)7.540.006* Long (≥ 3 months)252 (73.5%)84 (85.7%)336 (76.2%)Diagnostic interval Short (< 30 days)195 (56.9%)52 (53.1%)247 (56%)0.6640.415 Long (≥ 30 days)148 (43.1%)46 (46.9%)194 (44%)Data are presented as n (%)*Statistically significant at level of p < 0.05 Clinical characteristics of patients with cervical cancer diagnosed at UOG and FHCSH, Northwest Ethiopia Data are presented as n (%) *Statistically significant at level of p < 0.05 Tumor characteristics The majority of patients with cervical cancer [250 (56.7%)] were diagnosed at an advanced stage (stage III or IV). Stages IIB and IIA accounted for the largest proportions, with 115 (26.1%) and 134 (30.4%) cases, respectively. More than three-fourths [367 (83.2%)] had squamous cell carcinoma. Only 88 (20.2%) patients received surgical treatment at the study hospitals, and 15 (3.4%) had a history of radiotherapy through referral. Nearly half [214 (48.5%)] completed six cycles of chemotherapy. Over the 5-year follow-up, 98 (22.2%) deaths were recorded. Survival status differed significantly by stage at diagnosis (p < 0.0001), and treatment status; chemotherapy (p < 0.0001), and surgery (p = 0.02) (Table 3). Table 3Tumor characteristics of patients with cervical cancer diagnosed at UOG and FHCSH, Northwest EthiopiaCharacteristicsStatus at last contactTotalLog-rank testp-valueCensoredDeathStage at diagnosis I10 (2.9%)1 (1.0%)11 (2.5%)104.48 < 0.0001 II153 (44.6%)27 (27.6%)180 (40.8%) III167 (48.7%)62 (63.3%)229 (51.9%) IV13 (3.8%)8 (8.2%)21 (4.8%)Histological type Keratinizing squamous cell carcinoma118 (34.4%)35 (35.7%)153 (34.7%)4.110.1282 Non-keratinizing squamous cell carcinoma165 (48.1%)49 (50.0%)214 (48.5%) Adenocarcinoma60 (17.5%)14 (14.3%)74 (16.8%)Recurrence (n = 361) Yes12 (4.2%)4 (5.6%)16 (4.4%)2.860.091 No277 (95.8%)68 (94.4%)345 (95.6%)Chemotherapy Took ≥ 6 cycles182 (53.1%)33 (33.7%)215 (48.5%)19.26 < 0.0001 Took < 6 cycles161 (46.9%)65 (66.3%)226 (51.2%)History of surgery Yes80 (23.3%)9 (9.2%)88 (20%)9.77 0.020 No263 (76.7%)89 (90.8%)353 (80%)Data are presented as n (%)Boldface text: Statistically significant Tumor characteristics of patients with cervical cancer diagnosed at UOG and FHCSH, Northwest Ethiopia Data are presented as n (%) Boldface text: Statistically significant Mortality and survival The mortality incidence among patients with cervical cancer was 11.03 per 1,000 person-months (95% CI: 9.04–13.44), with a median survival time of 34 months. Overall survival at 2, 3, 4, and 5 years was 81.2%, 55.89%, 37.26%, and 18.63%, respectively. In the worst-case analysis, survival dropped to 65%, 39%, and 26% at 2, 3, and 4 years, respectively, with median survival reduced to 30 months in the sensitivity analysis. The 5-year survival rate was 57.61% for early-stage and 25.29% for advanced-stage patients. By stage, 5-year survival was 57.03% for stage II, 26.98% for stage III, and 9.04% for stage IV cervical cancer (Figs. 2, 3, 4 and 5). Survival status of patients with cervical cancer at UOG and FHCSH, Northwest Ethiopia, 2022 Survival status of patients with cervical cancer diagnosed at different stages of the disease at UOG and FHCSH, Northwest Ethiopia Four-year survival status of patients with cervical cancer based on chemotherapy status at UOG and FHCSH, Northwest Ethiopia, 2022 Survival status of patients with cervical cancer based on underlying comorbidity status at UOG and FHCSH, Northwest Ethiopia, 2022 Predictors of mortality In the multivariable Cox regression analysis, Age ≥ 50 years, HIV positivity, Travel distance > 65 km, no history of surgery, and advanced stage at diagnosis (Stage III and Stage IV) were identified as predictors of mortality following a cervical cancer diagnosis. Those cervical cancer patients whose age was greater than 50 years were 3.4 times at risk of death than those whose age was less than 50 years (AHR = 3.4, 95% CI: 1.80–6.48, p < 0.0001). HIV-positive patients were 2.9 times at risk of death than their counterparts (AHR = 2.9, 95% CI: 1.55–5.61, p = 0.001). Patients who traveled ≥ 65 km (more than the average distance from the oncology center) were 2 times at risk of death than those who had a traveling distance less than the average (AHR = 1.9, 95% CI: 1.28–3.08, p = 0.002). Cervical cancer patients who did not undergo surgical treatment were more likely to be at higher risk of death than patients who got surgical treatment (AHR = 2.8, 95% CI: 1.33–5.71, p = 0.006). Patients who were diagnosed with locally advanced disease (stage III) of the disease were 3.4 times at risk of death than those who were diagnosed with an early-stage disease (stage I&II) (AHR = 3.4, 95% CI: 1.95–5.79, p < 0.001). Patients who were diagnosed with metastatic disease (stage IV) were 6.7 times at risk of death than those who were diagnosed with an early-stage disease (AHR = 6.7, 95% CI: 3.35–13.58, p < 0.001). However, factors such as residence, comorbidity, patient delay, anemia and menopausal status were not significant predictors of mortality (P > 0.05) (Table 4). Table 4Bi-variable and multivariable cox regression analysis of predictors of mortality among patients with cervical cancer diagnosed at UOG and FHCSH, Northwest EthiopiaDeterminant factorsStatusCHR (95% CI)AHR (95% CI)p-valueDeathCensoredAge group < 50 years235 (68.5%)27 (27.6%)11< 0.0001* ≥ 50 years108 (31.5%)71 (72.4%)4.85 (3.11–7.57)3.4 (1.80–6.48)Residence Rural239 (69.7%)84 (85.7%)2.21 (1.26–3.90)1.5 (0.854–2.78)0.112 Urban104 (30.3%)14 (14.3%)11HIV status Positive309 (90.1%)84 (84.7%)1.48 (0.84–2.61)2.9 (1.55–5.61)0.001* Negative34 (9.9%)14 (14.3%)11Travel distance < 65 km186 (54.2%)39 (39.8%)110.002* ≥ 65 km157 (45.8%)59 (60.2%)2.09 (1.38–3.15)1.9 (1.28–3.08)Comorbidities Yes117 (34.1%)48 (49.0%)110.055 No226 (65.9%)50 (51.0%)1.93 (1.29–2.87)1.5 (0.99–2.30)Patient delay Short91 (26.5%)14 (14.3%)110.093 Long252 (73.5%)84 (85.7%)2.15 (1.22–3.79)1.7 (0.917–3.044)History of surgery Yes80 (23.3%)9 (9.2%)110.006* No263 (76.7%)89 (90.8%)2.81 (1.42–5.59)2.8 (1.33–5.71)Anemia Yes136 (39.7%)54 (55.1%)1.03 (1.01–2.25)1.0 (0.67–1.58)0.885 No207 (60.3%)44 (44.9%)11Menopausal status Pre-menopausal136 (39.7%)17 (17.3%)110.211 Menopausal207 (60.3%)81 (82.7%)3.48 (2.05–5.89)1.6 (0. 75–3.59)Stage at diagnosis Stage I&II174 (50.7%)17 (17.3%)11 Stage III167 (48.7%)62 (63.3%)3.8(2.14, 6.76))3.4 (1.95–5.79)< 0.001* Stage IV13 (3.8%)8 (8.2%)6.3 (2.29, 17.33)6.7 (3.35–13.58)< 0.001*Data are presented as n (%)CHR, crude hazard ratio; AHR, adjusted hazard ratio*Statistically significant association Bi-variable and multivariable cox regression analysis of predictors of mortality among patients with cervical cancer diagnosed at UOG and FHCSH, Northwest Ethiopia Data are presented as n (%) CHR, crude hazard ratio; AHR, adjusted hazard ratio *Statistically significant association Discussion Cervical cancer is the second leading cause of cancer death among African women. Although it is highly preventable, mortality remains higher in low-resource settings than in high-income countries for various reasons [14, 30, 31]. Approximately 85% of cervical cancer deaths occur in low- and middle-income countries, creating a significant economic burden [30]. This study assessed overall survival and predictors of mortality among patients with cervical cancer in Northwest Ethiopia. The mean age at diagnosis in our study was 49 years, lower than the global average of 53 years [31] but consistent with findings from Tikur Anbesa Hospital Radiotherapy Center, where the average age at diagnosis was 48 years [32]. The overall survival rate of patients at 52 months of follow-up was 18.63%, which is lower than findings from Tikur Anbesa Radiotherapy Center (38.62% and 28.4%) [21, 22] studies from other African countries (33.1%), and evidence from Tanzania (26%) [32, 33]. This variation may be due to limited access to radiotherapy services in our study settings—only 4.3% of patients had a history of receiving radiotherapy—which significantly impacts survival. Differences were also observed when compared with studies conducted at Felege Hiwot and University of Gondar hospitals, likely due to variations in how survival time was measured and differences in the analytical methods used [34, 35]. However, our finding is consistent with a pooled systematic review and meta-analysis in Ethiopia, which reported a pooled mortality rate of 16.39% among patients with cervical cancer [36]. After controlling for potential confounding factors, the multivariable Cox regression analysis revealed that age of ≥ 50 years, HIV-positive status, travel distance of ≥ 65 km, no history of surgical treatment, history of precancerous cervical lesions, and advanced stage at diagnosis were significant predictors of mortality among patients with cervical cancer. In our study, more than half (56.7%) of patients with cervical cancer presented at an advanced stage (III or IV). This finding is consistent with studies from Sub-Saharan Africa, including a meta-analysis in Ethiopia [19, 36]. This is largely due to the lack of screening and long patient and diagnostic delays, which are linked to poor access to health care and low health-seeking behavior. Patients with cervical cancer in Ethiopia often face prolonged health-seeking and diagnostic intervals [28]. Patients who received surgical treatment had better survival than those who did not. This finding aligns with other studies showing that increasing adherence to recommended treatment options improves survival among patients with cervical cancer [30, 33, 34, 37]. Surgery is often essential for managing early-stage cervical cancer and can significantly improve survival. The absence of surgery may indicate more advanced disease or reflect patient preferences and systemic barriers. This result is also consistent with a systematic review and meta-analysis in Ethiopia, which confirmed the effectiveness of surgery with adjuvant therapy in improving survival [36]. A similar finding was reported in China, where surgery prolonged survival in stage IA–IIA patients [38]. Interventions aimed at raising patient awareness about the benefits of surgery and addressing logistical barriers to accessing surgical services are crucial for improving survival. Older women had poorer survival rates in our study. Patients over 50 years of age were approximately three times more likely to die than those under 50 years of age. This suggests that 5-year survival is higher among younger patients with cervical cancer and declines with increasing age. The association between age at diagnosis and survival has been reported in studies from Ethiopia and other African countries, where an age of over 50 years was identified as an independent predictor of death among patients with cervical cancer [14, 22, 39, 40]. This might be due to older women being more likely to have comorbidities that accelerate disease progression. In our study, survival was lower among patients diagnosed at an advanced stage. Those diagnosed with stage III and IV were 3.4 and 6.7 times more likely to die than patients diagnosed with an early stages (stage I&II) respectively. There was also a significant difference in 5-year survival across stages: 57.03% for stage II, 26.98% for stage III, and 9.04% for stage IV, according to the log-rank test. This finding is supported by other studies [14, 21, 33, 36]. As disease stage advances, the likelihood of tumor metastasis increases, which worsens survival because survival is independently associated with the number of regional organs involved [15, 41]. Once metastasized, the disease damages vital organs, reducing life expectancy. Early-stage cervical cancer is treatable and has a much higher survival rate [19, 22, 42]. In Ethiopia, although some promising initiatives have begun, cervical cancer screening remains limited, and few centers provide regular screening despite the high prevalence of precancerous lesions [43, 44]. As a result, most women present at advanced stages. Government and non-governmental organizations need to work together to raise awareness about the disease and the importance of early detection. Our study also found that patients who had cervical cancer with HIV infection had poorer survival rates. HIV-positive patients were 2.9 times more likely to die than their HIV-negative counterparts. Similar findings have been reported in other studies from Ethiopia [22, 39]. This may be because HIV-positive patients have compromised immune systems, which can accelerate disease progression and reduce treatment effectiveness. This finding underscores the need for integrated care that addresses both HIV and cervical cancer, improves screening access for HIV-positive women, and promotes education on regular follow-ups. Our results are consistent with systematic reviews and meta-analyses from East Africa, where HIV status and comorbidities were the most frequently cited prognostic factors for survival [45]. Another important finding in this study was that patients who traveled long distances to the oncology centers were twice as likely to die and had lower survival rates than those living closer. This may be because patients from distant areas face poor access to oncology services, delayed diagnosis, and treatment initiation due to financial constraints or lack of transportation, making them more likely to present at advanced stages with metastasized disease. This geographic disparity highlights the need for strategies to improve access for rural and remote populations through decentralizing or expanding oncology centers and implementing community-based outreach programs. In general, compared with middle- and high-income countries [46, 47], the overall survival rate of patients in our study setting was significantly lower. It was also lower than the estimated global 5-year net survival for cervical cancer (42.1%), as well as findings from Bhutan (55.4%) [42] and Brazil (84%) [41]. This low survival rate may reflect disparities in health care infrastructure, including the effectiveness of cancer care systems in those countries and the absence of radiotherapy services in our study hospitals. In Sub-Saharan Africa, including Ethiopia, two-thirds of women never receive cancer-directed therapy despite having potentially curable disease, leading to poor survival outcomes. This highlights the urgent need for greater investment in radiotherapy, chemotherapy, and surgical treatment options [30]. Our study has certain limitations. As this was a retrospective study based on hospital records and supplementary phone interviews, potential recall bias and incomplete documentation cannot be entirely excluded. Since the data on patients medical records did not originally collected for research purposes, some variables were incomplete. Patients were excluded because of missing medical charts—possibly misplaced—or because they were lost to follow-up before starting treatment or had incomplete information, which may have affected our results. These excluded patients may have had worse prognoses than those included in the study. The absence of radiotherapy services in the study hospitals likely had the greatest impact on patient survival, although some patients who could afford it were referred to the national radiotherapy center. However, many did not receive the therapy, likely due to financial barriers, long travel distances, prolonged waiting times, loss of hope, or preference for traditional treatments. Because cause-of- death information was unavailable in most records, all-cause mortality was used for survival analysis. This may have led to over- or under-estimation of cancer-specific outcomes. Despite these limitations, this study provides valuable early data on survival status and predictors of mortality among cervical cancer patients in Northern Ethiopia. Moreover, the findings offer important insights for other low- and middle-income countries (LMICs) facing similar challenges. By identifying key prognostic factors influencing survival, this study underscores the urgent need for strengthening early detection programs, improving access to comprehensive cancer care, and enhancing health system capacity in comparable settings. The results may therefore serve as a reference point for policymakers and healthcare planners in other low-resource regions seeking to design context-appropriate strategies to improve cervical cancer outcomes. Conclusion The overall survival rate of patients with cervical cancer in our study was low. Key predictors of reduced survival were advanced stage at diagnosis, age of ≥ 50 years, HIV positivity, travel distance of ≥ 65 km, and absence of surgical treatment. There is an urgent need to expand cervical cancer screening with accurate diagnostic methods and to ensure timely, affordable, and high-quality diagnostic and treatment services. Expanding radiotherapy services in the region is crucial to reducing mortality, along with increasing access to surgery and chemotherapy at oncology centers. To better reach high-risk groups—such as older women and women living with HIV—the government should strengthen regular cervical cancer screening programs at all health care facilities as part of global elimination initiatives. The recent update of Ethiopia’s National Cervical Cancer Screening Guideline is a promising step, and improved survival is expected as more screening becomes available. Further prospective follow-up studies are needed to explore mortality predictors across diverse populations and multiple centers, informing targeted intervention strategies to address the unique challenges of different demographic groups. Supplementary Information Supplementary Material 1. Acknowledgements The authors would like to acknowledge the data collectors, supervisors, and staff of the oncology unit and record office at the study hospitals. We also thank the hospital administrators and medical directors for permitting us to conduct the study. Abbreviations Human Immunodeficiency Virus Felege Hiwot Comprehensive Specialized Hospital University of Gondar Hospital Authors’ contributions AT, SG, and EK contributed to the conceptualization and design of the study. AT, SG, AA, and EK were involved in proposal writing, project administration, analysis, and visualization. AT and SG drafted the original manuscript. AT, SG, AA, and EK contributed to the writing, analysis, and final manuscript review. All authors read and approved the final version of the manuscript. Funding This study was supported by the Else Kröner-Fresenius Foundation, Grant No. 2018_HA31SP. The project on which this publication is based was also partially funded by the German Federal Ministry of Research, Technology, and Space (BMFTR) 01KA2220B to the RHISSA Program for the NORA Consortium. Additional support was provided through a grant from Hospital Partnership via Deutsche Gesellschaft für Internationale Zusammenarbeit, funded by the Ministry for Economic Cooperation and Development (ID 81307397). This research was funded in whole or in part by the Science for Africa Foundation through the Developing Excellence in Leadership, Training and Science in Africa (DELTAS Africa) program with support from Wellcome Trust and the UK Foreign, Commonwealth & Development Office, and is part of the EDCTP2 program supported by the European Union. The funder has no role in the conceptualization, design, data collection, analysis, decision to publish, or preparation of the manuscript. Data availability The data sets used in this study are available from the corresponding author upon reasonable request. Declarations Ethics approval and consent to participate Ethical approval was obtained from the Research Ethics Committee of Debre Tabor University (Ref. No. DTU/RP/917/13). Due to the retrospective nature of the data collection from patient medical charts, institutional permission was obtained from each of the selected hospitals after a detail explanation of the purpose of the study to extract patient’s charts and contact the patients through phone call. Individual oral informed consent was obtained by phone from living patients, while consent for deceased patients were obtained from their relatives (husbands, mother father or children over the age of 18 years old). Confidentiality of patient information was maintained. All research methods followed the relevant guidelines and regulations. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Eva Kantelhardt and Sefonias Getachew share senior authorship equally. References Associated Data Supplementary Materials Supplementary Material 1. Data Availability Statement The data sets used in this study are available from the corresponding author upon reasonable request.